Deployment of Herbicide Ballistic Technology (HBT) in Hawaii: A case study on eliminating incipient invasive species targets with extreme prejudice
James Leary, Assistant Professor
Learning Outcomes:
• Uniqueness of Hawaii
• Biology and life history traits of a
tropical invasive plant species
• Concepts in aerial surveillance with
manned helicopter operations
• GIS analyses of operations data
• Bioeconomics of controlling an invasive
species
:
1,400 native plants in 5M yr
30 canoe plants in 2K yr
>10,000 introduced plants in 200 yr
Invasion History of Hawaii
• 90% of native species are endemic
• >10,000 introduced plants
• ~200 disruptive invasive species ecosystems and natural resources
Hawaii the Endangered/Invasive Species Capital of the World
Life History Traits of
Miconia (Miconia calvescens DC)
• An autogamous (self-fertile) species
• High fecundity
• Edible to broad range of frugivores
• Long dispersal range
• Persistent seed viability
• Germination in heavy shade
Meyer 1998, Murphy et al. 2008
Photo: F. Starr
East Maui Watershed (eastmauiwatershed.org):150,000 acres from sea level to 10,023′ a.s.l.
Critical habitat for over 100 threatened and endangered species
Annually produces 60 billion gallons of fresh surface water
(Largest single source of water in the State of Hawaii)
Miconia purposely introduced
as botanical specimen (founder
population)
Invasion History of Miconia in the East Maui Watershed
ca. 1970
Invasion History of Miconia in the East Maui Watershed
2011 (n=981,630)
Core infestation >2000 acres
Invasion spread across 30% of
the EMW
Herbicide Ballistic Technology (HBT) targeting Miconia (Miconia calvescens)
Photo: J. Atwood
Herbicide Ballistic Technology (HBT)
Problem: Incipient populations of miconia are colonizing remote sections of class I
watersheds on Maui. Average slope of terrain >50% >>> AREAS INACCESSIBLE TO GROUND
MANAGEMENT
Concept: Encapsulated 0.68 caliber herbicide-filled projectiles pneumatically delivered to
treat isolated plant populations >>> LONG DISTANCE ACCURACY (20-30 m) WITH FULL TILT
TRAJECTORY (0-90O)
Objective: Detect and eliminate incipient populations
0.68 caliber soft gel projectiles
encapsulating 200 mg triclopyr
HBT SD heliops
HBT operations conducted with a Hughes 500D helicopter
Payload Operator “Applicator” portside behind Pilot with shared 8-12 o’clock field of view (FOV)
Navigator front-starboard for 12-3 o’clock FOV and hazard awareness
Contour surveillance:
Low alt agl (≤30 m) at slow search speeds (10 knots)
Operational flight time 80-90 min per cycle
Scan rate is ~0.5 sec pixel-1 (100 m2)
Invasion History of Miconia in the East Maui Watershed
1991-2011 (n=981,630)
HBT 2012-2018 (n=25,956; yellow)
y = 39.588x + 83.657R² = 0.6266
0
100
200
300
400
500
600
0 2 4 6 8
sec
ha-1
Target ha-1
Variable costs of HBT operations determined by target density
$24.90 ha-1 and $20.71 tgt-1
• Search effort is a linear function of target density with search effort (EFT; y-intercept)
equal to 83 sec ha-1 = $24.90 and target engage equal to 40 sec tgt-1 = $12.00
• Herbicide use rate is also a linear function of target density with average dose (slope)
equal to 4.28 g ae (~22 projectiles) = $6.71
y = 4.2753xR² = 0.8126
0
10
20
30
40
0 2 4 6 8
Gra
ms
ae h
a-1Target ha-1
Search Effort Herbicide Use Rate
Propagule Dispersal and Persistence Kernels of Miconia
• Near analyses calculating the distance and time interval of progeny to maternal source.
• Dispersal kernels are leptokurtic with 99% of progeny within 587 m of the maternal source
and rare long-distance events out to 1640 m.
• Seed bank depletion 90% in <10 yrs; extinction estimated between 20-30 years
• These estimates are highly congruent to dispersal estimates in AU (Fletcher and Westcott
2013) and seed bank longevity in Tahiti (Meyer 2010)
Dispersal Recruitment
0
0.1
0.2
0.3
0 600 1200 1800
Pro
gen
y P
rop
ort
ion
Dispersal Range (m)
0.0
0.2
0.4
0.6
0.8
1.0
0 5 10 15 20
Pro
gen
y P
rop
ort
ion
Years after Reproduction
Spatial Impact and Future Cost of an Incipient Miconia Colonizing New Range
• Maximum dispersal distance creates an impact area >2000 acres
• >90% of total cost searching for 1% of targets
• The highest recruit densities cost >$300 pixel-1; effective search is $0.42 pixel-1
• The total estimated cost $42K to eradicate a new incipient population
0.01 tgt
y = $20.71x + $0.42
0
90
180
270
360
0481216
Co
st (
$ U
SD)
Targets Pixel-1
Dispersal kernel Density cost spectrum
• Only ~12% of impacted area (grey) surveyed (i.e., protected)
• >50% of mature plants detected on the first intervention (reconnaissance)
• Time between interventions is 2x longer for mature detection i.e., 180 days
Search freq.ReconnaissanceSearch effort
Management to Achieve Containment
2012-2018 We’ve invested over $799K, conducting 157 HBT operations:
Eliminating 25,956 incipient targets Protecting 17,433 ha of watershed Avoiding $1.1 B in future management costs
2012-2018 We’ve invested over $799K, conducting 157 HBT operations:
Eliminating 294 incipient MATURE targets Impacting 14,670 ha of watershed Adding min. $12.2 M in future management costs
98.9% effective!?!
Review of concepts:
• Knowledge of phenology, fecundity
and dispersal enhances our capacity
to measure impact on the landscape
and project long-term strategies
• GPS records of the area searched and
for every target treated offers the
most accurate cost assessment
• Operational cost is density dependent
• The most economical form of
management is searching and not
finding
Thank you very much.
Questions?